Scared? You are meant to be. Everything you have read about the US election has been cast in a way to keep you frozen in your seat until the show is over. It is the way we journalists do things. You thought Mitt Romney had a shot at becoming president four years ago? Ha! Got you! And judging by the jumpy stock markets and the spike in the price of gold, it looks like we’ve done it again: people who should know better appear to think that Donald Trump might win on Nov. 8.
Every reporter is taught that there is no story without tension. That is why every electoral contest feels like a nail-biter, even when the odds heavily favour one contestant. News outlets have kept us engaged in Trump vs. Hillary (Clinton) by emphasizing national polls over prediction models. The S&P 500 Index was on an eight-day losing streak, apparently because Trump caught Clinton in a series of nationwide and state polls. In a country as rigidly partisan as the US, it should surprise no one that the Democratic and Republican candidates would attract roughly equal amounts of core support. What matters is where those supporters live, and who stands the greater chance of persuading independents to vote for them. Traditional polls offer little in this regard. They might say something about the national mood, but they say much less about who stands the greatest chance of winning the 270 electoral college votes needed to become president. Jim Messina, who was Obama’s campaign manager, says traditional polls — the ones based on telephone surveys of a few hundred people — are useless. Better are the prediction models that combine polls with other data to put odds on the outcome of elections. There are lots of these and they have good success rates. But you are far more likely to have read recently that Clinton and Trump are tied than you are to have read that Trump has less than a 15% chance of becoming president, according to the New York Times's elections model. Larry Sabato, a political science professor at the University of Virginia who has been predicting the electoral college for years, says Clinton likely already has secured all the states she needs to win the White House, while Trump has no obvious path to victory.
Continue reading at the Centre for International Governance Innovation ...